---
base_model: deepseek-ai/deepseek-math-7b-base
datasets:
- zfz1/my_preference_gsm8k_deepseek
library_name: peft
license: other
tags:
- alignment-handbook
- trl
- orpo
- generated_from_trainer
model-index:
- name: deepseek-8b-orpo-lora
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/thuzfz1/huggingface/runs/zs68qfsi)
# deepseek-8b-orpo-lora

This model is a fine-tuned version of [deepseek-ai/deepseek-math-7b-base](https://huggingface.co/deepseek-ai/deepseek-math-7b-base) on the zfz1/my_preference_gsm8k_deepseek dataset.
It achieves the following results on the evaluation set:
- Loss: 1.1456
- Rewards/chosen: -0.0726
- Rewards/rejected: -0.0757
- Rewards/accuracies: 0.5647
- Rewards/margins: 0.0031
- Logps/rejected: -0.7569
- Logps/chosen: -0.7262
- Logits/rejected: 30.3288
- Logits/chosen: 30.4427
- Nll Loss: 1.0719
- Log Odds Ratio: -0.6909
- Log Odds Chosen: 0.0553

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 3e-06
- train_batch_size: 16
- eval_batch_size: 16
- seed: 43
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 2

### Training results



### Framework versions

- PEFT 0.11.1
- Transformers 4.42.3
- Pytorch 2.1.2
- Datasets 2.20.0
- Tokenizers 0.19.1